Track Connections at posit::conf (2023-2024)

Kamada–Kawai Forced-Directed Network analysis showing how conference tracks are related through shared words in talk titles.

TidyTuesday
Data Visualization
R Programming
2025
Exploring the interconnected topics at posit::conf through network analysis. This visualization reveals how different conference tracks are linked through common terminology in talk titles, highlighting the relationships between various R programming and data science themes across 2023-2024 conferences.
Author

Steven Ponce

Published

January 13, 2025

Figure 1: A Kamada-Kawai network visualization of posit::conf tracks (2023-2024), where tracks are nodes connected by gray edges. Thicker edges indicate more shared words between track titles, and larger blue nodes show tracks with more connections. The network reveals clusters of related topics and central tracks that bridge different conference themes.

Steps to Create this Graphic

1. Load Packages & Setup

Show code
## 1. LOAD PACKAGES & SETUP ----
suppressPackageStartupMessages({
if (!require("pacman")) install.packages("pacman")
pacman::p_load(
    tidyverse,      # Easily Install and Load the 'Tidyverse'
    ggtext,         # Improved Text Rendering Support for 'ggplot2'
    showtext,       # Using Fonts More Easily in R Graphs
    janitor,        # Simple Tools for Examining and Cleaning Dirty Data
    skimr,          # Compact and Flexible Summaries of Data
    scales,         # Scale Functions for Visualization
    glue,           # Interpreted String Literals
    here,           # A Simpler Way to Find Your Files
    tidytext,       # Text Mining using 'dplyr', 'ggplot2', and Other Tidy Tools 
    ggraph,         # An Implementation of Grammar of Graphics for Graphs and Networks 
    igraph,         # Network Analysis and Visualization 
    tidygraph       # A Tidy API for Graph Manipulation
    #withr           # Run Code 'With' Temporarily Modified Global State
)

})

# Source utility functions
suppressMessages(source(here::here("R/utils/fonts.R")))
source(here::here("R/utils/social_icons.R"))
source(here::here("R/utils/image_utils.R"))
source(here::here("R/themes/base_theme.R"))

2. Read in the Data

Show code
tt <- tidytuesdayR::tt_load(2025, week = 02) 

conf2023 <- tt$conf2023 |> clean_names()
conf2024 <- tt$conf2024 |> clean_names()

tidytuesdayR::readme(tt)
rm(tt)

3. Examine the Data

Show code
glimpse(conf2023)
skim(conf2023)

glimpse(conf2024)
skim(conf2024)

4. Tidy Data

Show code
# Prepare 2023 data
conf2023_clean <- conf2023 |>
    select(
        speaker_name,
        title = session_title,
        description = session_abstract,
        track = block_track_title,
        session_type,
        speaker_affiliation,
        session_date,
        session_start,
        session_length
    ) |>
    mutate(
        year = 2023,
        has_video = FALSE
    )

# Prepare 2024 data
conf2024_clean <- conf2024 |>
    select(
        speaker_name,
        title = talk_title,
        description,
        track,
        yt_url
    ) |>
    mutate(
        year = 2024,
        has_video = TRUE,
        session_type = case_when(
            str_to_lower(track) == "keynote" ~ "keynote",
            TRUE ~ "regular"
        ),
        speaker_affiliation = NA_character_,
        session_date = NA,
        session_start = NA,
        session_length = NA
    )

# Combine datasets
conf_combined <- bind_rows(conf2023_clean, conf2024_clean) 

### |-  plot data ----

# Create topic similarity network
title_similarity <- conf_combined |>
    # Split titles into individual words
    unnest_tokens(word, title) |>
    
    # Remove common stop words (e.g., "the", "and", "in")
    anti_join(stop_words) |>
    
    # Clean up words:
    # Remove numbers and single-letter words
    filter(!str_detect(word, "^[0-9]+$"),
           str_length(word) > 1) |>
    
    # Count word occurrences per track
    count(track, word) |>
    
    # Focus on meaningful patterns:
    group_by(track) |>
    # Keep words that appear at least twice
    filter(n >= 2) |>
    # Take top 8 most frequent words per track
    slice_max(n, n = 8) |>
    ungroup()

# Create network edges by finding pairs of tracks that share common words
edges <- title_similarity |>
    # Group by each unique word
    group_by(word) |>
    
    # Keep only words that appear in more than one track
    filter(n() > 1) |>
    
    # For each word group, create pairs of tracks that share the word
    summarize(
        combinations = list(data.frame(
            # First track in each pair (taking first row of combinations)
            X1 = combn(track, 2)[1,],
            # Second track in each pair (taking second row of combinations) 
            X2 = combn(track, 2)[2,]
        ))
    ) |>
    
    # Convert the list of combinations into rows
    unnest(combinations) |>
    
    # Count how many words each pair of tracks have in common
    # This creates the 'weight' of the connection between tracks
    count(X1, X2, name = "weight")

# Create network nodes
nodes <- tibble(
    name = unique(c(edges$X1, edges$X2)),
    type = "track"
)

# Create network graph using tidygraph
graph <- tbl_graph(
    nodes = nodes,
    edges = edges,
    directed = FALSE
) |>
    # Add degree centrality using mutate
    mutate(
        degree = centrality_degree()
    )

5. Visualization Parameters

Show code
### |-  plot aesthetics ----
# Get base colors with custom palette
colors <- get_theme_colors(palette = c("gray50", "#297ACC"))

### |-  titles and caption ----
title_text <- str_glue("Track Connections at posit::conf (2023-2024)")

subtitle_text <- str_glue(
    "__Kamada–Kawai__ Forced-Directed Network analysis showing how conference tracks are related<br>
    through shared words in talk titles.<br><br>
    __Node size__ corresponds to __node degree__ (the number of connections to other tracks),<br>
    __edge thickness__ shows the number of shared words between tracks."
)

# Create caption
caption_text <- create_social_caption(
    tt_year = 2025,
    tt_week = 02,
    source_text = "posit::conf attendee portal 2023-2024"
)

### |-  fonts ----
setup_fonts()
fonts <- get_font_families()

### |-  plot theme ----

# Start with base theme
base_theme <- create_base_theme(colors)

# Add weekly-specific theme elements
weekly_theme <- extend_weekly_theme(
    base_theme,
    theme(
        # Weekly-specific modifications
        legend.box = "vertical",          # Stack legends vertically
        # legend.position = "top",
        legend.position      = c(0.95, 1.28),    # x=1, y=1 puts it in the upper-right
        legend.justification = c(1, 1),          # Anchor the legend’s top-right corner
        legend.box.margin  = margin(b = 15),
        legend.spacing     = unit(0.2, "cm"),
        legend.box.spacing = unit(0.2, "cm"),
        legend.key.size    = unit(0.8, "lines"),
        legend.text        = element_text(size = 9),
        legend.title       = element_text(size = 10, face = "bold"),
        panel.grid.major   = element_blank(),
        panel.grid.minor   = element_blank()
    )
)

# Set theme
theme_set(weekly_theme)

6. Plot

Show code
### |-  Plot ----

# Set seed for reproducibility
set.seed(123)  

# Create layout using tidygraph functions
graph_laid_out <- graph |>
    activate(nodes) |>
    create_layout(layout = "kk")

# ggraph call
p <- ggraph(graph_laid_out) +
    # Geom
    geom_edge_link(
        aes(edge_alpha = weight, 
            edge_width = weight),
        edge_color = colors$palette[1],
        show.legend = TRUE,
        edge_linetype = "solid",
        alpha = 0.5,    
        lineend = "round"
    ) +
    geom_node_point(
        aes(size = degree),
        color = colors$palette[2],
        alpha = 0.8
    ) +
    geom_node_label(
        aes(label = str_wrap(name, 20)),
        repel        = TRUE,
        fill = alpha("white", 0.8),  
        label.size   = 0,     # remove border
        label.padding = unit(0.15, "lines"),
        size         = 3.0,
        family       = fonts$text,
        fontface     = "bold",
        color        = colors$text,
        # Additional ggrepel arguments:
        box.padding  = 0.4,        # Increase if labels overlap too much
        point.padding = 0.3,       # Space between node and label
        force        = 1.0,        # Higher = stronger repel
        force_pull   = 0.1,        # Pull label toward or away from point
        max.overlaps = Inf
    ) +
    # Scales
    scale_edge_width(range = c(0.5, 2.5)) +
    scale_size(range = c(3, 8)) +
    scale_edge_alpha(range = c(0.2, 0.8)) +
    
    # Labs
    labs(
        title = title_text,
        subtitle = subtitle_text,
        caption = caption_text,
        edge_alpha = "Shared Words",   
        edge_width = "Connection Strength",   
        size = "Node Degree" 
    ) +
    # Combine similar legends
    guides(
        edge_alpha = guide_legend(
            title = "Connection Strength",
            override.aes = list(alpha = 0.6)
        ),
        edge_width = "none",  # Hide duplicate legend
        size = guide_legend(
            title = "Node Degree",
            override.aes = list(alpha = 0.8)
        ) 
    ) +
    # Theme
    theme(
        plot.title = element_text(
            size   = rel(2.6),
            family = fonts$title,
            face   = "bold",
            color  = colors$title,
            lineheight = 1.1,
            margin = margin(t = 5, b = 5)
        ),
        plot.subtitle = element_markdown(
            size   = rel(1.1),
            family = fonts$subtitle,
            color  = colors$subtitle,
            lineheight = 1.2,
            margin = margin(t = 5, b = 15)
        ),
        plot.caption = element_markdown(
            size   = rel(0.65),
            family = fonts$caption,
            color  = colors$caption,
            hjust  = 0.5,
            margin = margin(t = 10)
        )
    )

7. Save

Show code
### |-  plot image ----  

save_plot(p, type = "tidytuesday", 
          year = 2025, week = 02, width = 14, height = 10)

8. Session Info

R version 4.4.1 (2024-06-14 ucrt)
Platform: x86_64-w64-mingw32/x64
Running under: Windows 11 x64 (build 22631)

Matrix products: default


locale:
[1] LC_COLLATE=English_United States.utf8 
[2] LC_CTYPE=English_United States.utf8   
[3] LC_MONETARY=English_United States.utf8
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.utf8    

time zone: America/New_York
tzcode source: internal

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

other attached packages:
 [1] tidygraph_1.3.1 igraph_2.1.1    ggraph_2.2.1    tidytext_0.4.2 
 [5] here_1.0.1      glue_1.8.0      scales_1.3.0    skimr_2.1.5    
 [9] janitor_2.2.0   showtext_0.9-7  showtextdb_3.0  sysfonts_0.8.9 
[13] ggtext_0.1.2    lubridate_1.9.3 forcats_1.0.0   stringr_1.5.1  
[17] dplyr_1.1.4     purrr_1.0.2     readr_2.1.5     tidyr_1.3.1    
[21] tibble_3.2.1    ggplot2_3.5.1   tidyverse_2.0.0 pacman_0.5.1   

loaded via a namespace (and not attached):
 [1] tidyselect_1.2.1   viridisLite_0.4.2  farver_2.1.2       viridis_0.6.5     
 [5] fastmap_1.2.0      gh_1.4.1           tweenr_2.0.3       janeaustenr_1.0.0 
 [9] digest_0.6.37      timechange_0.3.0   lifecycle_1.0.4    tokenizers_0.3.0  
[13] magrittr_2.0.3     compiler_4.4.0     rlang_1.1.4        tools_4.4.0       
[17] utf8_1.2.4         yaml_2.3.10        knitr_1.49         labeling_0.4.3    
[21] graphlayouts_1.2.0 htmlwidgets_1.6.4  bit_4.5.0          curl_6.0.0        
[25] xml2_1.3.6         repr_1.1.7         tidytuesdayR_1.1.2 withr_3.0.2       
[29] grid_4.4.0         polyclip_1.10-7    fansi_1.0.6        colorspace_2.1-1  
[33] gitcreds_0.1.2     MASS_7.3-60.2      cli_3.6.3          crayon_1.5.3      
[37] rmarkdown_2.29     ragg_1.3.3         generics_0.1.3     rstudioapi_0.17.1 
[41] tzdb_0.4.0         commonmark_1.9.2   cachem_1.1.0       ggforce_0.4.2     
[45] parallel_4.4.0     base64enc_0.1-3    vctrs_0.6.5        Matrix_1.7-0      
[49] jsonlite_1.8.9     hms_1.1.3          bit64_4.5.2        ggrepel_0.9.6     
[53] magick_2.8.5       systemfonts_1.1.0  codetools_0.2-20   stringi_1.8.4     
[57] gtable_0.3.6       munsell_0.5.1      pillar_1.9.0       rappdirs_0.3.3    
[61] htmltools_0.5.8.1  R6_2.5.1           httr2_1.0.6        textshaping_0.4.0 
[65] rprojroot_2.0.4    vroom_1.6.5        evaluate_1.0.1     lattice_0.22-6    
[69] markdown_1.13      SnowballC_0.7.1    gridtext_0.1.5     memoise_2.0.1     
[73] snakecase_0.11.1   renv_1.0.3         Rcpp_1.0.13-1      gridExtra_2.3     
[77] xfun_0.49          pkgconfig_2.0.3

9. GitHub Repository

The complete code for this analysis is available in tt_2025_02.qmd.

For the full repository, click here.

10. References

  1. Data Sources:
Back to top